habanalabs/gaudi: add FW alive event support

In order for driver to be aware of process or thread crashes inside
GAUDI's CPU, we introduce a new event which contains all relevant
information. Upon event reception, driver will dump information and
will reset the device.
Signed-off-by: Ofir Bitton <obitton@habana.ai>
Reviewed-by: Oded Gabbay <ogabbay@kernel.org>
Signed-off-by: Oded Gabbay <ogabbay@kernel.org>

drivers/misc/habanalabs/gaudi/gaudi.c

@@ -7451,6 +7451,16 @@ static void gaudi_print_out_of_sync_info(struct hl_device *hdev,
 			sync_err->pi, sync_err->ci, q->pi, atomic_read(&q->ci));
 }
 
+static void gaudi_print_fw_alive_info(struct hl_device *hdev,
+					struct hl_eq_fw_alive *fw_alive)
+{
+	dev_err(hdev->dev,
+		"FW alive report: severity=%s, process_id=%u, thread_id=%u, uptime=%llu seconds\n",
+		(fw_alive->severity == FW_ALIVE_SEVERITY_MINOR) ?
+		"Minor" : "Critical", fw_alive->process_id,
+		fw_alive->thread_id, fw_alive->uptime_seconds);
+}
+
 static int gaudi_soft_reset_late_init(struct hl_device *hdev)
 {
 	struct gaudi_device *gaudi = hdev->asic_specific;
@@ -7902,6 +7912,11 @@ static void gaudi_handle_eqe(struct hl_device *hdev,
 		gaudi_print_out_of_sync_info(hdev, &eq_entry->pkt_sync_err);
 		goto reset_device;
 
+	case GAUDI_EVENT_FW_ALIVE_S:
+		gaudi_print_irq_info(hdev, event_type, false);
+		gaudi_print_fw_alive_info(hdev, &eq_entry->fw_alive);
+		goto reset_device;
+
 	default:
 		dev_err(hdev->dev, "Received invalid H/W interrupt %d\n",
 				event_type);

drivers/misc/habanalabs/include/common/cpucp_if.h

@@ -84,6 +84,20 @@ struct hl_eq_sm_sei_data {
 	__u8 pad[3];
 };
 
+enum hl_fw_alive_severity {
+	FW_ALIVE_SEVERITY_MINOR,
+	FW_ALIVE_SEVERITY_CRITICAL
+};
+
+struct hl_eq_fw_alive {
+	__le64 uptime_seconds;
+	__le32 process_id;
+	__le32 thread_id;
+	/* enum hl_fw_alive_severity */
+	__u8 severity;
+	__u8 pad[7];
+};
+
 struct hl_eq_entry {
 	struct hl_eq_header hdr;
 	union {
@@ -91,6 +105,7 @@ struct hl_eq_entry {
 		struct hl_eq_hbm_ecc_data hbm_ecc_data;
 		struct hl_eq_sm_sei_data sm_sei_data;
 		struct cpucp_pkt_sync_err pkt_sync_err;
+		struct hl_eq_fw_alive fw_alive;
 		__le64 data[7];
 	};
 };

drivers/misc/habanalabs/include/gaudi/gaudi_async_events.h

@@ -303,6 +303,7 @@ enum gaudi_async_event_id {
 	GAUDI_EVENT_NIC3_QP1 = 619,
 	GAUDI_EVENT_NIC4_QP0 = 620,
 	GAUDI_EVENT_NIC4_QP1 = 621,
+	GAUDI_EVENT_FW_ALIVE_S = 645,
 	GAUDI_EVENT_DEV_RESET_REQ = 646,
 	GAUDI_EVENT_PKT_QUEUE_OUT_SYNC = 647,
 	GAUDI_EVENT_FIX_POWER_ENV_S = 658,

drivers/misc/habanalabs/include/gaudi/gaudi_async_ids_map_extended.h

@@ -669,7 +669,7 @@ static struct gaudi_async_events_ids_map gaudi_irq_map_table[] = {
 	{ .fc_id = 642, .cpu_id = 491, .valid = 0, .name = "" },
 	{ .fc_id = 643, .cpu_id = 492, .valid = 0, .name = "" },
 	{ .fc_id = 644, .cpu_id = 493, .valid = 0, .name = "" },
-	{ .fc_id = 645, .cpu_id = 494, .valid = 0, .name = "" },
+	{ .fc_id = 645, .cpu_id = 494, .valid = 1, .name = "FW_ALIVE_S" },
 	{ .fc_id = 646, .cpu_id = 495, .valid = 1, .name = "DEV_RESET_REQ" },
 	{ .fc_id = 647, .cpu_id = 496, .valid = 1,
 		.name = "PKT_QUEUE_OUT_SYNC" },